* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download Quantum Spacetimes and Finite N Effects in 4D Super Yang
Bra–ket notation wikipedia , lookup
Measurement in quantum mechanics wikipedia , lookup
Quantum decoherence wikipedia , lookup
Basil Hiley wikipedia , lookup
Asymptotic safety in quantum gravity wikipedia , lookup
Relativistic quantum mechanics wikipedia , lookup
Quantum dot wikipedia , lookup
Probability amplitude wikipedia , lookup
Hydrogen atom wikipedia , lookup
Copenhagen interpretation wikipedia , lookup
Quantum entanglement wikipedia , lookup
Density matrix wikipedia , lookup
Coherent states wikipedia , lookup
Quantum fiction wikipedia , lookup
Bell's theorem wikipedia , lookup
Quantum electrodynamics wikipedia , lookup
Quantum computing wikipedia , lookup
Quantum chromodynamics wikipedia , lookup
Path integral formulation wikipedia , lookup
Many-worlds interpretation wikipedia , lookup
Quantum teleportation wikipedia , lookup
Introduction to gauge theory wikipedia , lookup
Renormalization wikipedia , lookup
Renormalization group wikipedia , lookup
Quantum field theory wikipedia , lookup
Quantum machine learning wikipedia , lookup
Quantum key distribution wikipedia , lookup
EPR paradox wikipedia , lookup
Orchestrated objective reduction wikipedia , lookup
Interpretations of quantum mechanics wikipedia , lookup
AdS/CFT correspondence wikipedia , lookup
Quantum state wikipedia , lookup
Topological quantum field theory wikipedia , lookup
Canonical quantum gravity wikipedia , lookup
Scalar field theory wikipedia , lookup
Hidden variable theory wikipedia , lookup
History of quantum field theory wikipedia , lookup
Canonical quantization wikipedia , lookup
Brown Het-1190 arXiv:hep-th/9907145v1 17 Jul 1999 Quantum Spacetimes and Finite N Effects in 4D Super Yang-Mills Theories Pei-Ming Ho1 , Sanjaye Ramgoolam2 and Radu Tatar2 1 National Taiwan University, Taipei 10764, Taiwan, R.O.C. 2 Brown University Providence, RI 02912 [email protected], [email protected], [email protected] The truncation in the number of single-trace chiral primary operators of N = 4 SYM and its conjectured connection with gravity on quantum spacetimes are elaborated. The model of quantum spacetime we use is AdSq5 ×Sq5 for q a root of unity. The quantum sphere is defined as a homogeneous space with manifest SUq (3) symmetry, but as anticipated from the field theory correspondence, we show that there is a hidden SOq (6) symmetry in the constrution. We also study some properties of quantum space quotients as candidate models for the quantum spacetime relevant for some Zn quiver quotients of the N = 4 theory which break SUSY to N = 2. We find various qualitative agreements between the proposed models and the properties of the corresponding finite N gauge theories. 9/99 1. Introduction and summary The Maldacena duality [1,2,3,4] gives a relation between type IIB string theory on AdS5 × S 5 and the N = 4 superconformal four dimensional super Yang-Mills. The gauge group of the field theory is SU (N ) when the flux through S 5 is N . In the case of large N and large effective coupling, Maldacena’s conjecture relates the corresponding field theory and the classical supergravity. Finite N effects contain important information about the qualitative novelties of quantum gravity compared to classical gravity. It was proposed in [5] that the quantum corrections in the AdS × S background have the effect of deforming spacetime to a non-commutative manifold. The concrete model studied there was AdS3 × S 3 where the group structure of the manifold allowed a simple non-commutative candidate by using quantum groups. An important part of the evidence was a quantum group interpretation of the cutoff on single particle chiral primaries, first studied under the heading of “stringy exclusion principle” in [6]. Here we develop the same line of argument to understand analogous cutoffs in the spectrum of chiral primaries of N = 4 super Yang-Mills. The cutoffs originate from the fact that the U (N ) invariants of the form trΦl , where Φ is a matrix in the adjoint representation, are not independent for all values of l and the set of independent invariants truncates at N . Since the Yang-Mills theory has an SO(6) global symmetry, we get an SO(6) covariant cutoff on the chiral primaries. In the large N limit, the chiral primaries are matched on the gravitational side with modes coming form KK reduction on S 5 . To understand the cutoff at finite N , we postulate that the S 5 is deformed to a quantum sphere in the following way: S 5 has a description as a coset space SU (3)/SU (2) which generalizes to the q-deformed case as SUq (3)/SUq (2). This construction is shown to have SOq (6) hidden symmetry by mapping it to another construction of a q-sphere based on [7], thus explaining why KK reduction on it gives a truncated set of reps. of SO(6). We then discuss N = 2 theories obtained by taking Zn quotients of U (N n) theories with N = 4 supersymmetry, which are dual to gravity on AdS5 × S 5 /Zn . We discuss the cutoffs in the spectrum of chiral primaries in the quotient theories. The main point is that, if we ignore states coming from twisted sectors associated with non-trivial reps. of Zn , the cutoffs occur at N n as in the parent theory. To find candidates for the quantum space dual of the quiver theory we identify appropriate automorphisms of the dual theory, which are used to quotient the quantum Sq5 to give a space with SUq (2) × Uq (1) symmetry. When twisted sectors are taken into account, the cutoff in some chiral primaries charged 1 under the U (1) happens not at N n but at N . We begin a discussion of the quantum space explanation of this change of cutoffs. This requires the description of the Sq5 as an S 1 fibration over a q-deformed ball, which is acted upon by the quotient. 2. Truncation of generating chiral primary operators in N = 4 super Yang-Mills Consider the chiral primaries of this theory which are of the form : Ca1 a2 ···al tr(Φa1 · · · Φal ) (2.1) where the C are traceless symmetric tensors of SO(6). These symmetric tensors can be decomposed under SU (3) and contain the symmetric rep. of SU (3) corresponding to a Young tableau with one row of length l. The polynomials corresponding to this Young tableau are Ci1 i2 ···il tr(Φi1 Φi2 · · · Φil ), where the C are symmetric, and the ik are indices running from 1 to 3 and the corresponding scalars are complex. It is useful in the discussion of cutoffs to decompose the invariant polynomials in reps. of SU (3). In order to obtain the decomposition of symmetric representations of SO(6) into representations of SU (3), we can use the isomorphism between SO(6) and SU (4) under which the vector representation of SO(6) goes into the antisymmetric representation of SU (4), and each symmetric traceless representation of SO(6) goes into the (0, k, 0) representation of SU (4). The branching rules for the (0, k, 0) representations of SU (4) into representations of SU (3) are, for example, For k = 1 6 → 3 ⊕ 3̄, For k = 2 20 → 6 ⊕ 6̄ ⊕ 8, For k = 3 ¯ + 15 + 15. ¯ 50 → 10 + 10 (2.2) So the vector representation of SO(6) gives a fundamental and an anti-fundamental representation corresponding to tr(Φi ) and tr(Φ∗i ) (i=1,2,3), which include the chiral primary operator of dimension 1. The symmetric traceless representation 20 gives the following operators : tr(Φ(i1 Φi2 ) ) (the 6 representation), tr(Φ∗(i1 Φ∗i2 ) ) (the 6̄ representation), tr(Φ(i1 Φ∗i2 ) ) (the 8 representation). These include the chiral primary operator of dimension 2. The representation 50 of SO(6) gives the following : tr(Φ(i1 Φi2 Φi3 ) ) (the 10 representation), its conjugate one (involving the complex conjugate fields and correspond¯ representation), tr(Φ(i1 Φ∗i2 Φ∗i3 ) ) (the 15 representation) and its complex ing to the 10 2 ¯ representation). These contain the chiral primary of dimension 3. We conjugate (the 15 can continue the discussion to show that each symmetric traceless rep. of SO(6) contains a chiral primary belonging to a rep. of SU (3) associated to a symmetric Young tableau. The chiral primary operators are not all independent at finite N . Consider first the operators which look like tr(ΦN+1 ) when the gauge group is U (N ). If the generators of the Lie algebra are T a , the chiral primary operator is written as tr(Φa T a )N+1 = Φa1 · · · ΦaN +1 tr(T a1 · · · T aN +1 ). But tr(T a1 · · · T aN +1 ) is just the CN+1 Casimir operator of U (N ) and we know that U (N ) has only N independent Casimir operators so CN+1 is not independent and can be written in terms of lower Casimir operators. Therefore, the conclusion is that tr(ΦN+1 ) can be written in terms of tr(ΦN ), tr(ΦN−1 ) and so on. Thus tr(ΦN+1 ) does not describe a single particle state. The conclusion is that for the group U (N ) we have a truncation on the chiral primary operators such that the highest power is N. For instance, Φ can be one of the three Φi ’s. The rest of the symmetric polynomials like tr(Φi1 Φj2 Φk3 ) can also be decomposed since they can be obtained from the chiral primary by action of SO(6). In fact for any given dimension, there is only one short representation, so that operators of the form tr(F ΦΦ · · ·) (2.3) can also be shown to be decomposable as they are obtained from the chiral primary by action of the SUSY operators. The result is that we have a truncation on the short representations for the gauge group U (N ), the maximal symmetric traceless representation of SO(6) being the one with N boxes in the Young tableau. This will allow us to identify a candidate quantum sphere relevant for the spacetime understanding of finite N effects. After describing some preliminaries on quantum groups we will describe the relevant quantum space in section 3.3 3. Non-commutative spacetime Natural non-commutative candidates for AdS5 × S 5 are obtained by deforming the coset struture of the spaces involved using quantum groups. We will be concerned with some detailed properties of the q-deformed S 5 in this paper, which are relevant for the P6 2 truncation in KK modes. The unit five sphere is the space i=1 xi = 1 where xi are coordinates in R6 . SO(6) acts transitively on the solutions of this equation. A point, say (0, 0, 0, 0, 0, 1) is left fixed by SO(5). This allows an identification of the sphere with the 3 coset SO(6)/SO(5). It is posible to consider R6 as a complex space C3 with coordinates P2 z0 = x1 + ix2 , z1 = x3 + ix4 , z2 = x5 + ix6 and the sphere becomes a surface j=0 |zj |2 = 1 in C3 . In this case the sphere is seen as a coset SU (3)/SU (2). The latter coset space struture allows a simple quantum group generalization. 3.1. Preliminaries of quantum groups The standard q-deformation of quantum groups is given in [7] . For a matrix T i j of a quantum group, the commutation relations among the matrix elements are given by R̂12 T1 T2 = T1 T2 R̂12 . (3.1) This is a shorthand of the following ij k kl R̂kl T m T l n = T i k T j l R̂mn . (3.2) The matrix elements T i j in the fundamental rep. generate the algebra of functions on the quantum group. The matrix T has an inverse T −1 given by the antipode. The antipode is an automorphism S of this algebra such that S(T i j ) = (T −1 )ij . For SLq (N ; C), the R̂ matrix is given by ij R̂kl = δli δkj (1 + (q − 1)δ ij ) + λδki δlj θ(j − i), (3.3) where θ(j − i) = 1 if j > i and θ(j − i) = 0 otherwise. In [7] a ∗-anti-involution is given for SLq (N ; C). With respect to this ∗-anti-involution, a real form of SLq (N ; C) can be defined by (T i j )∗ = (T −1 )ji . For q being a phase, the real form SLq (N ) can be defined by T ∗ = T ; and for q being real, the real form SUq (N ) can be defined. (We require that the ∗ of a complex number be its complex conjugation, so e.g. q ∗ = q −1 if q is a phase.) In this paper what we need is SUq (N ), but there is no ∗-anti-involution for this purpose when q is a phase. It turns out that the appropriate ∗-structure is an involution instead of anti-involution. (An involution does not reverse the ordering of a product and an anti-involution does.) Let g i j = q i δji , (3.4) where i, j = 0, 1, · · · , N − 1, then we define T † = g −1 T −1 g, 4 (3.5) where T † is the transpose of T ∗ . One can check that this definition gives a ∗-involution. First, because S(S(T )) = g 2 T g −2 , (T ∗ )∗ = T . Secondly, (3.1) is invariant under the action of ∗. To check this, one can use the following identities ij kl R̂kl = R̂ij , (3.6) R̂(q −1 )12 = R̂(q)−1 21 , (3.7) g1−1 g2−1 R̂12 g1 g2 = R̂12 . (3.8) 3.2. The quantum sphere To begin we define the quantum complex plane CN q which has the symmetry group SUq (N ) acting on it. The algebra of functions on CN q is generated by the coordinates zi , i = 0, 1, · · · , N − 1, which satisfy the following commutation relations z1 z2 = q −1 z1 z2 R̂12 . (3.9) kl zi zj = q −1 zk zl R̂ij . (3.10) More explicitly, it is The coordinates z transform under an SUq (N ) matrix T as z → zT. (3.11) We let all zk ’s to commute with all T i j ’s. Due to (3.1) , the relations (3.9) is preserved by this transformation. (Note that T i j commutes with zk for all i, j, k.) The complex conjugation of z is defined as a ∗-involution. Let z̄ i = zi∗ . The ∗ of (3.9) is z̄1 z̄2 = q −1 R̂21 z̄1 z̄2 , (3.12) which is covariant under the transformation z̄ → T † z̄, as the ∗ of (3.11) . To complete the definition of the algebra on CN q , we also need to define how z commutes with z̄. Let z̄1 z1 = q −1 z2 g2 R̂12 g1−1 z̄2 , (3.13) ij l z̄ . z̄ i zk = q −1+j−k zj R̂kl (3.14) which means 5 It is covariant under the action of SUq (N ). The coefficient of q −1 on the right hand side of (3.13) is chosen such that the radius squared r 2 = zg z̄ (3.15) is a central element in the algebra. One can also check that r 2 is real: (r 2 )∗ = r 2 . Since r 2 commutes with everything else, we can define a new algebra by the algebra of z, z̄, modulo the condition r 2 = 1. This is the algebra of functions on the quantum sphere Sq2N−1 . It can be identified with the quantum sphere defined as SUq (N )/SUq (N − 1) [8] . Explicitly, the commutation relations of z, z̄ are zi zj = qzj zi , i < j, (3.16) zi z̄ j = qz̄ j zi , i 6= j, (3.17) z̄ i z̄ j = q −1 z̄ j z̄ i , i < j, X zi z̄ i = z̄ i zi − q −1 λ q j−i zj z̄ j . (3.18) (3.19) j>i Another natural candidate for the q-deformed sphere is the SOq (N )-covariant quantum Euclidean space RN q modulo the unit radius condition. The quantum group SOq (N ; C) is defined by (3.1) with a different R̂ matrix, which also has the properties (3.6) and (3.7). In addition, one has C1 C2 R̂12 C1 C2 = R̂21 , (3.20) i where C i j = δN+1−j . For q being a phase, the real form SOq (n, n) or SOq (n, n + 1) can be defined by T ∗ = T for a ∗-anti-involution. For real q, the real form SOq (N ; R) exists with respect i to the ∗-anti-involution T ∗ = GT G−1 , where Gi j = q i δN+1−j . We need SOq (N ; R) for q being a phase, so again we can only define it with respect to a ∗-involution. We find the appropriate ∗-involution to be given by T ∗ = CT C. (3.21) Incidentally, if one wants to define SOq (2n, 2m), we can generalize the above to T ∗ = i ĈT Ĉ −1 , where Ĉji = ǫi δN+1−j (N = 2(n + m)) with m of the ǫ’s equal to −1 and the rest to 1. This includes SO(4, 2) which is of interest in defining the deformation of the 6 AdS part. The ∗-involution in the corresponding universal enveloping algebra is recently given in [9], where some useful information about unitary representations when q is a root of unity is also given. In the following we will concentrate on SOq (N ; R), which will be denoted as simply SOq (N ). The SOq (N )-covariant algebra of functions on the quantum Euclidean space is defined by [7] x1 x2 = q −1 x1 x2 R̂12 + κR2 G, (3.22) where κ = (1 − q −2 )/(1 + q N−2 ) and R2 = xt Gx is the radius squared. The transformation of SOq (N ) on x is x → xT . The ∗-involution compatible with (3.21) is x∗ = xC. (3.23) Since we have demanded all algebras to have the ∗-involution, it follows that there is a symmetry of q → q −1 if we simultaneously take x → x∗ , T → T ∗ etc. It is therefore equivalent to say that we have Sq5 or Sq5−1 . It can be explicitly checked that the algebra of C3q is the same as the algebra of R6q via the identification zi = xi+1 and z̄ i = x6−i for i = 0, 1, 2, and then R2 = (q 2 + q −2 )r 2 . Therefore the three definitions of Sq5 are actually equivalent: SUq (3)/SUq (2) = C3q /(r 2 = 1) = R6q /(R2 = q 2 + q −2 ). In the first two models of Sq5 , the action of SUq (3) is manifest. In the third model the action of SOq (6) is manifest. This suggests that SUq (3) can be realized as a subgroup of SOq (6). This is indeed the case. Given a 3 × 3 SUq (3) matrix t, we can define a special 6 ×6 SOq (6) matrix T by Tij = tij for i, j = 1, 2, 3, Tij = t∗(7−i)(7−j) for i, j = 4, 5, 6, and all other elements Tij = 0. 3.3. Quantum sphere for U (N ) N = 4 SYM Now that we have established the SOq (6) symmetry of the quantum sphere, we use the fact that KK reduction on this space will give a family of reps. of SOq (6). In Sec. 3.5 we will show that SOq (6) can be identified with SUq (4). For q being a root of unity, the reps. of SUq (4) contain indecomposable reps. which form an ideal under tensor products. After quotienting these out, one is left with a set of standard reps. with a truncation. iπ The length of the first row of the Young tableau cannot exceed k for q = e k+4 . This fact is familiar from 2D WZW models [10,11] and has been studied in detail for Uq SU (2) in [12][13]. This allows us to identify k ∼ N to get a quantum sphere which gives a KaluzaKlein reduction which agrees with the spectrum of chiral primaries discussed in section 2. 7 3.4. Zn automorphisms and symetries of quantum quotient spaces We now discuss some sub-algebras relevant for the quantum space analog of the N = 2 quotient theories which will be discussed in the next section. The quantum group SUq (3) has SUq (2) as a subgroup. Given a 2 × 2 SUq (2) matrix t, we can define a 3 × 3 SUq (3) matrix T by Tij = tij for i, j = 1, 2, T33 = 1 and all other elements Tij = 0. The group Zn mentioned in previous sections can then be embedded in SUq (2) as a diagonal matrix diag(ω, ω −1 ), where ω n = 1. Classically, SO(6) has a maximal subgroup SU (2) × SU (2) × U (1). This is also true for SOq (6). The presence of a Uq (SU (2)) × Uq (SU (2)) × Uq (U (1)) subalgebra in Uq (SO(6)) or Uq (SU (4)) is clear from the definition of these algebras using the q-analog of the Chevalley-Serre basis [14]. From the point of view of the algbera of functions on SOq (6) we can describe a SUq (2) × SUq (2) × Uq (1) as follows. The first SUq (2) is Tmn = α γ 0 0 0 0 β δ 0 0 0 0 0 0 1 0 0 0 0 0 0 1 0 0 0 0 0 0 α −γ 0 0 0 . 0 −β δ (3.24) 0 b 0 0 0 0 1 0 0 d 0 0 0 −b 0 . 0 0 d (3.25) 0 0 0 . 0 0 1 (3.26) The second SUq (2) is Tmn = a 0 0 a 0 0 0 0 c 0 0 −c 0 0 1 0 0 0 The Uq (1) is Tmn = 1 0 0 0 0 0 0 0 1 0 0 Λ 0 0 0 0 0 0 0 0 0 Λ−1 0 0 0 0 0 0 1 0 These 6 × 6 matrices corresponding to the three subgroups SUq (2), Uq (1) and SUq (2) commute with one another, and they all satisfy the relation (3.1) for SOq (6). Note that in checking whether these matrices commute with one another, we should take any entry in a matrix to commute with any entry in another matrix. The reason for this is that if we take two matrices T and T ′ from a quantum group, for their product to satify the same 8 RTT relation (3.1) we should let all entries in T to commute with all entries in T ′ . (The functions on two copies of the group manifold is given in a tensor product.) The Zn symmetry we quotient by is a subalgebra of Uq (1) which is embedded as Tij = T(i+4)(j+4) = diag(ω, ω −1 ) for i, j = 1, 2, T33 = T44 = 1, and all other elements vanishing, for |ω| = 1. The commutant of the Zn action is SUq (2) × Uq (1), with the same q as before. This fact will be useful in a quantum space-time understanding of the relation between cutoffs on chiral primaries in an N = 4 U (N n) theory and its Zn quotient. 3.5. SUq (4) symmetry Since the universal enveloping algebra Uq (G) for a classical group G is completely determined by the Cartan matrix of G, and since the Cartan Matrix of SU (4) is the same as that for SO(6), SOq (6) is identical to SUq (4) ( up to global differences which do not affect the general sub-group structure). Therefore SOq (6) has a subgroup SUq (3) which is manifest in the SUq (4) description. While we arrived at the SUq (4) ∼ SOq (6) symmetry by explicitly mapping to an algebra which had the larger symmetry, we can also guess it by an indirect argument. A hint for the hidden SU (4)q comes from a consideration of KK reduction on SUq (3)/SUq (2). Suppose we are dimensionally reducing a scalar on the coset SU (3)/SU (2). We have to look for all reps. of SU (3) which contain a scalar of SU (2) [15]. We know that the reps. of SU (3) we have are precisely such that they combine into reps. of SU (4). To KK reduce on SUq (3)/SUq (2), we q-deform the rule above and look for reps. of SUq (3) which contain the scalar of SUq (2). It is very plausible that the reps. still combine into reps. of SUq (4) since the structure of the reps. at roots of unity remains the same as long as we stay within the cutoff. This can be proved by using the generalization of the Gelfand-Zetlin bases ( described for example in [16]), which exists because of some special properties of the branching rules in the sequence of subgroups SU (2) ⊂ SU (3) ⊂ SU (4). The GelfandZetlin bases have been generalized to roots of unity ( discussed for example in [17] ) so this should provide the proof that the desired q-generalization of the branching rules is correct. The matching between SUq (4) and SOq (6) can also be illustrated as follows. There is an SUq (3) subgroup of SUq (4) which acts on (z1 , z2 , z3 ). Let us represent it as a b c d e f g h p 0 0 0 9 0 0 . 0 1 (3.27) This corresponds to the SUq (3) embedded in SOq (6) which is described in a previous section. In addition to the SUq (2) subgroups in this SUq (3), there are three other ways of embedding SUq (2) in SUq (4): a 0 0 c 0 1 0 0 0 0 1 0 b 0 , 0 d 1 0 0 0 0 a 0 c 0 0 0 b , 1 0 0 d 1 0 0 0 0 1 0 0 0 0 0 0 , a b c d (3.28) where a, b, c, d are the four functions which constitute an SUq (2) matrix in the fundamental representation. These three SUq (2) subgroups all commute with one another. They are characterized by their being commuting with different SUq (2) subgroups of the SUq (3) mentioned above. It is now not difficult to guess what their correspondence in SOq (6) is. The corresponding SUq (2) matrices are: 1 0 0 0 a 0 0 0 a 0 c 0 0 0 −c 0 0 0 0 b 0 d 0 0 0 0 −b 0 d 0 0 0 0 , 0 0 1 a 0 0 c 0 0 0 0 1 0 0 a 0 0 0 0 0 −c b 0 0 d 0 0 0 0 0 0 1 0 0 0 −b , 0 0 d (3.29) and a 0 0 a 0 0 0 0 c 0 0 −c 0 0 1 0 0 0 0 0 0 1 0 0 b 0 0 −b 0 0 . 0 0 d 0 0 d (3.30) The matching above takes care of the 8 generators in the SUq (3) and 2 generators in each of the three SUq (2). Since there are a total of 15 generators in SUq (4) or SOq (6), there is still one generator left, the U (1) generator which can be represented as Λ 0 0 0 0 Λ 0 0 0 0 Λ 0 0 0 , 0 Λ−3 in SUq (4) and SOq (6) respectively. 10 Λ13×3 0 0 −1 Λ 13×3 (3.31) 4. Chiral primary operators for N = 2 quotient theories 4.1. Conformal field theory discussion Maldacena’s conjecture has been extended to the case of orbifolds. In order to preserve the conformal symmetry, we need to keep the AdS part untouched and to act with orbifold groups only on S 5 [18,19,20,21]. An N = 2 theory is obtained if we act with a Zn group on two out of three complex fields, one of them being left unchanged. The Zn quotienting is accompanied by a gauge transformation. ΩΦ1 Ω−1 = ωΦ1 , ΩΦ2 Ω−1 = ω −1 Φ2 , (4.1) ΩΦ3 Ω−1 = Φ3 , ΩDA Ω−1 = DA , where DA is the covariant derivative. The Φ’s are N n × N n matrices. Ω can be chosen to be diag(1, ω −1, ω −2 · · · ω −(n−1) ). After taking the quotient the gauge group becomes SU (N )⊗n , with the surviving gauge fields being diagonal N × N blocks, and the bosonic matter content is : Φ1 = Φ2 = Φ3 = (1) 0 ··· 0 ··· , 0 ··· .. . . . . ··· ··· ··· , ··· .. . 0 ··· 0 ··· , 0 ··· .. . . . . 0 Q1 0 0 (2) 0 0 Q1 0 (3) 0 0 0 Q1 .. .. .. .. . . . . 0 0 0 0 (1) Q2 0 0 0 (2) 0 Q2 0 0 (3) 0 0 Q2 0 .. .. .. .. . . . . (1) Φ̂3 0 0 .. . 0 (2) Φ̂3 0 .. . 0 0 (3) Φ̂3 .. . (i) (4.2) (i) where Q1 are fields in the (Ni , N̄i+1 ) representation, and Q2 are in the (N̄i , Ni+1 ) repre(i) (i) sentation. The surviving global symmetry is SU (2)R ×U (1)R ×Zn . The pair (Q1 , (Q2 )∗ ) (i) is a doublet of SU (2)R and uncharged under U (1)R , while Φ3 are singlets under this 11 SU (2)R but have charge 1 under the U (1). The SU (2)R has a U (1) subgroup, under which (i) (i) the Q1 and Q2 have charge 1. The Zn acts as cyclic permutations on the n factors of (i) (i) SU (N ) and on the i index of Q1 , Q2 , Φi3 . Geometrically, the symmetry SU (2)×U (1)×Zn can be understood, by describing S 5 /Zn as a fibration of S 2 × S 1 over S 2 /Zn . Out of the chiral primaries of the N = 4 theory, those giving non-trivial operators l are of the form trΦl3 , tr(Φ1 Φ2 )l and tr(Φ1 Φ2 Φm 3 ) . As in the discussion of cutoffs for the U (N n) theory we can write traces of powers of these fields in terms of products of traces when the power exceeds N n. For example this leads to tr(Φ3 )Nn+1 = tr(Φ3 )Nn tr(Φ3 ) + · · · , (4.3) where the · · · stands for other terms involving other splittings of the (N n + 1)’th power. We can rewrite this as follows tr(Φ3 )Nn+1 = X (i) tr(Φ̂3 )Nn X (j) tr(Φ̂3 ) + · · · . (4.4) j i Note that the splitting involves factors which are separately ZN invariant. The same holds for other operators, e.g. tr(Φ1 Φ2 )Nn+1 . Another kind of splitting occurs if we allow the factors to come from twisted sectors. This splitting will happen at a different value of the powers, i.e at N + 1 rather than N n + 1. It uses the fact that (i) (i) (i) tr(Φ̂3 )N+1 = tr(Φ̂3 )N tr(Φ̂3 ) + · · · , and it leads to tr(Φ3 ) N+1 = n X (i) (i) tr(Φ̂3 )N tr(Φ̂3 ) + · · · . (4.5) (4.6) i=1 Now the factors are not separately invariant. They come from twisted sectors. Similar equations can be written for the other operators, for example, tr(Φ1 Φ2 ) N+1 = n X (i) (i) (i) (i) tr(Φ̂1 Φ̂2 )N tr(Φ̂1 Φ̂2 ) + · · · (4.7) i=1 However as discussed in [21] the factors appearing on the right hand side in this expression are not chiral primaries because they appear as derivatives of a superpotential. 12 4.2. Quantum space explanation of the cutoffs We saw in the previous subsection that if we start with a gauge theory with gauge group U (N n) and take the Zn quotient, we get a gauge theory with a product of U (N ) gauge groups. Restricting attention to Zn invariant operators, i.e the untwisted sector, we find that the cutoff stays at N n. We expect on the gravity side that a discussion ignoring the twisted sectors can be simply reproduced by studying a quotient of the q-sphere. We found indeed in the previous section that after quotienting Sq5 by a Zn element inside the SOq (6), we are left with a surviving SUq (2) × Uq (1) with the same value of q that we started with. This gives an explanation of the fact that the cutoff ∼ N n stays at ∼ N n. 4.3. Comments on twisted sectors The quantum space explanation of the behaviour of the cutoffs when we include the twisted sectors is much more intricate. This is only a preliminary discussion. The twisted sector has to do with dimensional reduction on the singular cycles of D4 /Zn followed by the reduction on S 1 [21]. To understand the twisted sector chiral primaries from the gravity point of view in the large N limit one starts with a description of S 5 /Zn as S 1 fibred over B4 /Zn wher B4 is the 4-ball. Ten-dimensional gravity is reduced on the B4 /Zn and then the KK reduction on the S 1 is performed. The cohomology of the blown-up B4 /Zn space is used to determine the type of particles we get [21]. We can try to generalize this discussion to the case of Sq5 /Zn . The twisted sector states are localized in the Z1 , Z2 direction when the Zn quotient acts in these directions. The wavefunctions of the twisted sector states are arbitary functions of the phase of Z3 and localized at Z1 = Z2 = 0 in the case q = 1. A very intriguing property of the Sq5 we described is that Z1 = Z2 = 0 is not compatible with the algebra (3.16) - (3.19) . In a sense the origin of the ball is smoothed out. It would be very interesting if this result of noncommutativity provides an effective way to describe the resolution of the fixed point in a similar way as [22] , where it was shown that the instanton moduli space can be resolved by deforming the base space into a quantum space. Another possibility is that we could also have chosen to work with a Zn action on Z2 , Z3 . In that case there is a circle over Z2 = Z3 = 0 on Sq5 . So to describe the twisted sector states, we need to extend the algebra Sq5 /Zn by adding certain delta functions mutliplied with arbitrary powers of Z3 ( with unit norm ). The results from the field theory suggest that this algebra should admit a 13 consistent truncation which restricts the U (1) charge at order N , since we saw that with the cutoffs on single trace operators of the form trΦl3 happens at order N . It will be very interesting to see if the quantum space techniques can reproduce this field theory result. Some relevant techniques on quantum principal bundles may be found in [23,24] and on K-theory of quantum spaces in [25]. 5. Conclusions In this paper we have considered the AdS/CFT conjecture for finite N conformal theories. As opposed to the large N limit which relates the corresponding field theories and the classical supergravity, for finite N we need to consider quantum gravity. We have taken the point of view that the finite N effects can be captured by gravity on a q-deformed version of AdS5 × S 5 space time where q is a root of unity along the lines of [5]. We have discussed mainly the deformation of S 5 into Sq5 where q is a root of unity. iπ The KK reduction on this quantum sphere is truncated at N if q = e N +4 . This agrees with the cutoff on the chiral primaries of finite N conformal field theory obtained on the boundary of AdS5 . We then discussed the N = 2 quotient theories and the corresponding cutoff in the supergravity and field theory. By considering only the untwisted sectors the result is that the chiral primaries have the same cutoff as in the initial N = 4 theory. This result could be explained by considering a quotient of the quantum sphere. We also derived results in field theories regarding the cutoffs when twisted sectors are taken into account. We began a discussion of the corresponding quantum space picture. It would be interesting to extend the results of this paper for other orbifolds of S 5 , for six dimensional field theories obtained on the boundary of AdS7 or for three dimensional theories obtained on the boundary of AdS4 . Some evidence for non-commutative gravity in AdS7 × S 4 background has been obtained in [26] and connections to the quantum group approach will be interesting to explore. Acknowledgements : We are happy to acknowledge fruitful discussions with Antal Jevicki, Gilad Lifschytz, and Vipul Periwal. P.M.H. was supported in part by the National Science Council, Taiwan, R.O.C., and he thanks C.-S. Chu for discussions and the hospitality of the University of Neuchâtel, Switzerland, where part of this work was done. The work of R.T. and S.R. was supported by DOE grant DE-FG02/19ER40688-(Task A). S.R would like to thank the National Taiwan University for hospitality while part of this work was done. 14 References [1] J. Maldacena, The Large N Limit of Superconformal Field Theories and Supergravity, Adv.Theor.Math.Phys. 2 (1998) 231-252. [2] S. S. Gubser, I. R. Klebanov and A. M. Polyakov, Gauge Theory Corellators from Non-Critical String Theory, Phys.Lett. B428 (1998) 105-114. [3] E. Witten, Anti-de-Sitter space and holography, hepth-9802150. [4] O. Aharony, S. S. Gubser, J. Maldacena, H. Ooguri, Y. Oz, Large N Field Theories, String Theory and Gravity, hep-th/9905111. [5] A. Jevicki, S. Ramgoolam, Non-commutative gravity from the AdS/CFT correspondence, JHEP 9904 (1999) 032. [6] J. Maldacena, A. Strominger, AdS3 Black Holes and a Stringy Exclusion Principle, JHEP 9812 (1998) 005. [7] L. D. Faddeev, N. Yu. Reshetikhin, L. A. Takhtadzhyan, Quantization of Lie groups and Lie algebras, Leningrad Math. J. 1 (1990) 193. [8] L. L. Vaksman and Ya. S. Soibelman The Algebra of functions on the quantum group SU (n + 1) and odd-dimensional quantum spheres, Leningrad Math. J. Vol. 2 ( 1991 ), No. 5. [9] H. Steinacker, Unitary Representations of Noncompact Quantum Groups at Roots of Unity, math.QA/9907021; Unitary Representations and BRST Structure of the Quantum Anti-de Sitter Group at Roots of Unity, q-alg/9710016; Quantum Groups, Roots of Unity and Particles on Quantized Anti-de Sitter Space, hep-th/9705211; Finite Dimensional Unitary Representations of Quantum Anti-de Sitter Groups at Roots of Unity, q-alg/9611009. [10] V. Pasquier, H. Saleur,Common Structures Between Finite Systems and Conformal Field through Quantum Groups, Nucl.Phys. B330 (1990) 523. [11] P. Furlan, A. Ch. Ganchev, V. B. Petkova, Quantum Groups and Fusion Rule Multiplicities, Nucl. Phys. B343 (1990) 205. [12] G. Keller, Fusion Rules of Uq SL(2, C), q m = 1, Lett. Math. Phys. 21: 273, 1991. [13] N. Reshetikhin, V. G. Turaev, “Invariants of three manifolds via link polynomials and quantum groups”, Invent.Math.103:547-597,1991. [14] M. Jimbo, A q-analog of U(gl(N+1) Hecke Algebras and the Yang-Baxter Equation, Lett. Math. Phys. 11 (1986), 247 - 252. [15] A. Salam, J. Strathdee, On Kaluza-Klein Theory, Annals in Phys. 141, 316, 1982. [16] A. O. Barut and R. Raczka, Theory of Group Representations and Applications PWN, 1980. [17] V. Chari, A. Pressley, A guide to quantum groups, CUP, 1994. [18] S. Kachru, E. Silverstein, 4d Conformal Field Theories and Strings on Orbifolds, Phys.Rev.Lett. 80 (1998) 4855-4858. 15 [19] A. Lawrence, N. Nekrasov, C. Vafa, On Conformal Theories in Four Dimensions, Nucl.Phys. B533 (1998) 199-209. [20] Y. Oz, J. Terning, Orbifolds of AdS5 ×S 5 and 4d Conformal Field Theories, Nucl.Phys. B532 (1998) 163-180. [21] S. Gukov, Comments on N=2 AdS Orbifolds, Phys.Lett. B439 (1998) 23-28. [22] N. Nekrasov, A. Schwarz, Instantons on Noncommutative R4 and (2, 0) Superconformal Six Dimensional Theory, hep-th/9802068. [23] C.-S. Chu, P.-M. Ho, H. Steinacker, q-deformed Dirac Monopole With Arbitrary Charge, Z.Phys. C71 (1996) 171-177. [24] T. Brzezinski and S. Majid, Quantum Group Gauge Theory on Quantum Spaces Commun.Math.Phys. 157 (1993) 591. [25] A. Connes, Non-commutative geometry, Academic Press 1994 [26] M. Berkooz and H. Verlinde, Matrix Theory, AdS/CFT and Higgs-Coulomb Equivalence, hep-th/9907100. 16